Reinforcement Learning with Partial Parametric Model Knowledge
arXiv:2304.13223 · doi:10.1016/j.ifacol.2023.10.924
Abstract
We adapt reinforcement learning (RL) methods for continuous control to bridge the gap between complete ignorance and perfect knowledge of the environment. Our method, Partial Knowledge Least Squares Policy Iteration (PLSPI), takes inspiration from both model-free RL and model-based control. It uses incomplete information from a partial model and retains RL's data-driven adaption towards optimal performance. The linear quadratic regulator provides a case study; numerical experiments demonstrate the effectiveness and resulting benefits of the proposed method.
IFAC World Congress 2023
References in corpus (6)
- Proceedings of the 29th International Conference on Machine Learning (ICML-12)
- Parameter Space Noise for Exploration
- Deep Reinforcement Learning for Process Control: A Primer for Beginners
- Meta-Reinforcement Learning for the Tuning of PI Controllers: An Offline Approach
- Mastering Atari Games with Limited Data
- Finite-time Analysis of Approximate Policy Iteration for the Linear Quadratic Regulator